Method for recognizing illumination state of traffic lights, electronic device, and storage medium

Abstract

A vehicle-borne method for recognizing the illumination state of traffic lights even against a backlighting of strong sunlight or other light source obtains a first image of a set of traffic lights in a road traffic environment. A segmentation map is acquired by dividing a first region from the first image, and an illumination region in the segmentation map is extracted by marking RGB pixels in the region which are of a preset threshold in brightness according to a training model. A lit color of the set of traffic lights is recognized according to a position of the illumination region in the segmentation map. By utilizing the method, accuracy of recognition of illumination state of traffic lights is improved.

What is claimed is: 
     
       1. A method for recognizing traffic lights, comprising:
 acquiring a first image that comprises a set of traffic lights; 
 recognizing a lit color of the set of traffic lights in the first image; 
 in response that no lit color of the traffic lights is recognized, confirming that the first image is in a backlight scene; 
 when confirming that the first image is in the backlight scene, obtaining a segmentation map by dividing a first region from the first image, in which the first region is an image region of the set of traffic lights; 
 extracting an illumination region from the segmentation map, comprising: dividing the segmentation map into three regions equally, wherein each of the three regions corresponds to a signal light in the set of traffic lights; determining a brightness of each region of the three regions, and determining a region with the highest brightness among the three regions as the illumination region; 
 recognizing a lit color of the set of traffic lights according to a position of the illumination region in the segmentation map, comprising: acquiring a location of the illumination region and a location of each of the three regions; determining a region where the illumination region is located from the three regions as a target region, according to the location of the illumination region and the location of each of the three regions; outputting a lit color corresponding to the target region. 
 
     
     
       2. The method of  claim 1 , wherein extracting an illumination region from the segmentation map comprises:
 converting a red green blue (RGB) color gamut of pixels in the segmentation map to a hue, saturation, value (HSV) color gamut, and obtaining a converted segmentation map; 
 extracting the illumination region from the converted segmentation map using a preset threshold. 
 
     
     
       3. The method of  claim 2 , wherein extracting the illumination region from the converted segmentation map using a preset threshold comprises:
 acquiring a value of a V channel of each pixel in each of the three regions in the converted segmentation map; 
 determining pixels whose values of the V channel in each of the three regions are greater than the preset threshold as first pixels of the corresponding region; 
 determining a brightness of each of the three regions according to first pixels of each of the three regions, and determining the region with a highest brightness as the illumination region. 
 
     
     
       4. The method of  claim 3 , further comprising:
 acquiring values of the V channel of all pixels in the converted segmentation map; 
 obtaining the preset threshold by averaging the values. 
 
     
     
       5. The method of  claim 1 , wherein obtaining a segmentation map by dividing a first region from the first image comprises:
 obtaining the segmentation map by using a segmentation network to divide the first region from the first image. 
 
     
     
       6. An electronic device comprising:
 a processor; and 
 a storage device storing a plurality of instructions, which when executed by the processor, cause the processor to: 
 acquire a first image that comprises a set of traffic lights; 
 recognize a lit color of the set of traffic lights in the first image: 
 in response that no lit color of the traffic lights is recognized, confirm that the first image is in a backlight scene; 
 when confirming that the first image is in the backlight scene, obtain a segmentation map by dividing a first region from the first image, in which the first region is an image region of the set of traffic lights; 
 extract an illumination region from the segmentation map, comprising: dividing the segmentation map into three regions equally, wherein each of the three regions corresponds to a signal light in the set of traffic lights: determining a brightness of each region of the three regions, and determining a region with the highest brightness among the three regions as the illumination region; 
 recognize a lit color of the set of traffic lights according to a position of the illumination region in the segmentation map, comprising: acquiring a location of the illumination region and a location of each of the three regions; determining a region where the illumination region is located from the three regions as a target region, according to the location of the illumination region and the location of each of the three regions: outputting a lit color corresponding to the target region. 
 
     
     
       7. The electronic device of  claim 6 , wherein the processor is further caused to:
 convert a red green blue (RGB) color gamut of pixels in the segmentation map to a hue, saturation, value (HSV) color gamut, and obtain a converted segmentation map; 
 extract the illumination region from the converted segmentation map using a preset threshold. 
 
     
     
       8. The electronic device of  claim 7 , wherein the processor is further caused to:
 acquire a value of a V channel of each pixel in each of the three regions in the converted segmentation map; 
 determine pixels whose values of the V channel in each of the three regions are greater than the preset threshold as the first pixels of the corresponding region; 
 determine a brightness of each of the three regions according to first pixels of each of the three regions, and determining the region with a highest brightness as the illumination region. 
 
     
     
       9. The electronic device of  claim 8 , wherein the processor is further caused to:
 acquire values of the V channel of all pixels in the converted segmentation map; 
 obtain the preset threshold by averaging the values. 
 
     
     
       10. The electronic device of  claim 6 , wherein the processor is further caused to:
 obtain the segmentation map by using a segmentation network to divide the first region from the first image. 
 
     
     
       11. A non-transitory storage medium having stored thereon at least one computer-readable instructions that, when executed by a processor of an electronic device, causes the processor to perform a method for recognizing traffic lights, the method comprising:
 acquiring a first image that comprises a set of traffic lights; 
 recognizing a lit color of the set of traffic lights in the first image; 
 in response that no lit color of the traffic lights is recognized, confirming that the first image is in a backlight scene; 
 when confirming that the first image is in the backlight scene, obtaining a segmentation map by dividing a first region from the first image, in which the first region is an image region of the set of traffic lights; 
 extracting an illumination region from the segmentation map, comprising: dividing the segmentation map into three regions equally, wherein each of the three regions corresponds to a signal light in the set of traffic lights; determining a brightness of each region of the three regions, and determining a region with the highest brightness among the three regions as the illumination region; 
 recognizing a lit color of the set of traffic lights according to a position of the illumination region in the segmentation map, comprising: acquiring a location of the illumination region and a location of each of the three regions; determining a region where the illumination region is located from the three regions as a target region, according to the location of the illumination region and the location of each of the three regions; outputting a lit color corresponding to the target region. 
 
     
     
       12. The non-transitory storage medium of  claim 11 , wherein extracting an illumination region from the segmentation map comprises:
 converting a red green blue (RGB) color gamut of pixels in the segmentation map to a hue, saturation, value (HSV) color gamut, and obtaining a converted segmentation map; 
 extracting the illumination region from the converted segmentation map using a preset threshold. 
 
     
     
       13. The non-transitory storage medium of  claim 12 , wherein extracting the illumination region from the converted segmentation map using a preset threshold comprises:
 acquiring a value of a V channel of each pixel in each of the three regions in the converted segmentation map; 
 determining pixels whose values of the V channel in each of the three regions are greater than the preset threshold as the first pixels of the corresponding region; 
 determining a brightness of each of the three regions according to first pixels of each of the three regions, and determining the region with a highest brightness as the illumination region. 
 
     
     
       14. The non-transitory storage medium of  claim 13 , further comprising:
 acquiring values of the V channel of all pixels in the converted segmentation map; 
 obtaining the preset threshold by averaging the values. 
 
     
     
       15. The non-transitory storage medium of  claim 11 , wherein obtaining a segmentation map by dividing a first region from the first image comprises:
 obtaining the segmentation map by using a segmentation network to divide the first region from the first image.